- Email: [email protected]
Time domain nuclear magnetic resonance as a method to determine and characterize the water-binding capacity of whey protein microparticles
Food Hydrocolloids 54 (2016) 170e178
Contents lists available at ScienceDirect
Food Hydrocolloids journal homepage: www.elsevier.com/locate/foodhyd
Time domain nuclear magnetic resonance as a method to determine and characterize the water-binding capacity of whey protein microparticles Jorien P.C.M. Peters a, Frank J. Vergeldt b, Henk Van As b, Hannemieke Luyten c, Remko M. Boom b, Atze Jan van der Goot a, * a b c
Food Process Engineering Group, Wageningen University, PO Box 18, 6700 AA Wageningen, The Netherlands Laboratory of Biophysics and Wageningen NMR Centre, Wageningen University, Dreijenlaan 3, 6703 HA Wageningen, The Netherlands FrieslandCampina, Stationsplein 4, 3818 LE Amersfoort, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 July 2015 Received in revised form 24 September 2015 Accepted 30 September 2015 Available online 9 October 2015
Water-binding capacity (WBC) is commonly measured with a centrifugation method in which a sample is hydrated in excess water and the pellet weight after centrifugation defines the WBC. When a dispersion is being analyzed, here containing whey protein microparticles (MPs), the pellet consists of swollen particles and water between the particles. These two water domains in MP pellets were distinguished using time domain nuclear magnetic resonance (TD NMR). This distinction showed that an increase in WBC from 2 to 5.5 g water/g dry matter was mainly due to an increase in water between the MPs. Besides, it was found that TD NMR-measurements could be used to provide accurate values of the amount of water in both water domains in MP pellets. This makes TD NMR therefore a more accurate method to determine the WBC of the whole pellet than weighing the pellet after centrifugation. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Proteins Water-binding capacity Swelling Nuclear magnetic resonance Centrifugation Microstructure
1. Introduction An important functional property of proteins is their ability to bind water. Unfortunately, this property is not well defined; various definitions and terms are used in the literature (Fennema, 1996; Kneifel, Paquin, Abert, & Richard, 1991; Zayas, 1997). Here, we use the term water-binding capacity (WBC) to describe the ability of a protein sample present in excess water to bind water when subjected to an external force. Several methods can be used to determine the WBC of a protein sample, but commonly a centrifugation method is used in which the obtained pellet weight determines the WBC. In this method, the protein sample is first placed in excess water to hydrate, and then surplus water is separated from the proteins by centrifugation. The WBC is then calculated using the
* Corresponding author. E-mail address: [email protected] (A.J. van der Goot). http://dx.doi.org/10.1016/j.foodhyd.2015.09.031 0268-005X/© 2015 Elsevier Ltd. All rights reserved.
following equation:
WBC ðg water=g dry matterÞ ¼
weightpellet weightdry matter weightdry matter (1)
where weightpellet is the pellet weight after centrifugation and weightdry_matter is the weight of the dry matter of the particles. In this study, water-binding of whey protein microparticles (MPs) is investigated to better understand the contribution of swelling of MPs and the water between MPs on the WBC. Previous research with MPs (Peters, Luyten, Alting, Boom, & van der Goot, 2015) suggested that the amount of water bound by the pellet is not necessarily a measure of the amount of water bound by the MPs, because a certain amount of water is bound between the MPs. Therefore, more insight is required into the ratio of water inside and between MPs. Because the pellet weight obtained after
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
centrifugation is used for other dispersions as well to determine the WBC (Ahmed, Al-Jassar, & Thomas, 2015; Berghout, Boom, & van der Goot, 2014; Jain, Prakash, & Radha, 2015; Sun & Xiong, 2014; Yu, Ahmedna, & Goktepe, 2007), the results of this study will have broader applicability and consequences. Time domain nuclear magnetic resonance (TD NMR) or relaxometry seemed to be a promising method to gain more understanding of WBC for MP pellets (van Duynhoven, Voda, Witek, & €tz, & Weisser, Van As, 2010; Goetz & Koehler, 2005; Hinrichs, Go 2003; Hinrichs et al., 2004; Mitchell, Gladden, Chandrasekera, & Fordham, 2014). TD NMR is a non-destructive method and is used to determine, inter alia, the transverse relaxation time (T2) of water via the transverse relaxation of 1H protons (Mitchell et al., 2014). T2 or the transverse relaxation rate R2 (R2 ¼ 1/T2) of every water fraction provides insight into the degree of exchange of either water protons with protein protons or the exchange of water between water fractions such as bulk water with the immobile water fraction around the proteins (Mariette, 2006). Since several mechanisms, such as diffusive and chemical exchange, and crossrelaxation, affect the T2 at the same time, a direct interpretation of the TD NMR results is difficult (Mariette, 2006). However, these mechanisms also cause that the T2 of water in a protein sample is related to the local protein and water concentrations, and the nanoand microstructure (e.g. pores and capillaries) of the protein network (Mariette, 2006; Oakes, 1976a, 1976b). Because both the protein concentration and the protein network structure are different inside and between the MPs in a pellet (Peters et al., 2015), we hypothesize that TD NMR can be used to distinguish between these two water domains. Several studies on meat (Bertram, Andersen, & Karlsson, 2001; Bertram, Dønstrup, Karlsson, & Andersen, 2002; Brøndum et al., 2000; Straadt, Rasmussen, Andersen, & Bertram, 2007) investigated the applicability of TD NMR to measure and gain insight into the WBC and WHC (e.g. the ability of meat to hold water under gravitational force). A good correlation was found between T2 of the water fraction assigned to the weakest bound water in meat and its WBC or WHC. In addition, the amplitude of this water fraction had a good correlation with the WBC or WHC. Therefore, it was concluded that TD NMR is a suitable method to gain insight into the WBC and WHC of meat. In addition to meat, TD NMR was found to be useful to examine the WHC of fresh cheeses (Hinrichs et al., 2004), carrageenan gels and solutions, whey gels, and yoghurt (Hinrichs et al., 2003). Though, to our knowledge, pellets obtained after the frequently used method of centrifuging protein dispersions to measure the WBC of those proteins have not been investigated yet. In this study, several modified MPs with different WBCs were made. The WBC of these MPs was determined through centrifugation of a dispersion of MPs and determining the pellet weight. The MPs and the pellet formed after centrifugation were further characterized with light scattering and TD NMR. These results were related to the WBC of the pellets to better understand the relationship between the swelling capacity of the MPs, the amount of water between the MPs, and the WBC of the pellets of MPs. In addition, a method is shown how TD NMR can be used to determine the WBC of MP pellets. 2. Materials and methods 2.1. Materials Whey protein isolate (WPI; BiPRO, lot no. JE 034-7-440, Davisco Food International Inc., Le Sueur, MN) was used to produce whey protein microparticles (MPs), gels, and solutions. WPI has a reported protein content of 97.6% on a dry basis. The water-binding
171
capacity (WBC) of the MPs was changed by treatment with dithiothreitol (DTT), N-ethylmaleimide (NEM), genipin (all from SigmaeAldrich, Germany), HCl, or microbial Ca2þ-independent transglutaminase (Tgase; Activa WM; activity of 81e135 units g1 according to the manufacturer) derived from Streptoverticillium mobaraense (Barentz Raw Materials, the Netherlands). Isopropanol Emplura (Merck, Germany) was used as a carrier fluid in static light scattering measurements. Potassium acetate (CH3CO2K), potassium carbonate (K2CO3), and potassium sulfate (K2SO4) (synthesis grade; SigmaeAldrich, Germany) were used to make saturated salt solutions. Milli-Q water was used (resistivity of 18.2 MU cm at 25 C, total oxidizable carbon <10 ppb; Merck Millipore, France) in all experiments, unless stated otherwise. 2.2. Methods 2.2.1. Formation of standard whey protein microparticles Standard MPs were made as described in Peters et al. (2015). WPI was dissolved in water to obtain a 40% w/w solution. This solution was mixed using an overhead mixer at 100 rpm and room temperature for at least 2 h, and then mixed further with a magnetic stirrer at 4 C overnight. The solution was centrifuged at 1000 rpm and 20 C for 10 min. After removing the foam layer manually, the solution was mixed in a bowl mixer (type W50) connected to a Brabender Do-corder E330 (Brabender OHG, Duisburg, Germany). The mixer was heated with water at 95 C using a water bath. The mixer settings were as follows: 5 min at 0 rpm, 5 min at 5 rpm, and 40 min at 200 rpm. After mixing, the mixer was cooled with water from a 4 C water bath for approximately 5 min. The wet gel particles were taken out of the mixer and placed in an oven at 50 C for 2 days. The dried gels were milled with an ultracentrifuge mill (Retsch ZM 1000) equipped with a 24-tooth stainless steel rotor. An 80-mm sieve was attached to the rotor, which operated at 15,000 rpm for 120 s. The particles obtained after milling were the standard MPs. 2.2.2. Formation of whey protein microparticles with altered waterbinding capacities MPs with an altered WBC were made by adapting the standard procedure (Section 2.2.1). MPs treated with DTT, a combination of DTT and NEM, or genipin were made by adding 25 g of wet gel particles to a 50-mL solution containing the chemicals. Concentrations of 10, 20, or 40 mM DTT were used with or without 15 mM NEM, and 0.5, 1, 2, or 4 mM of genipin. The dispersions were incubated and mixed at 20 C for 24 h. Part of the wet gel particles with DTT were washed after incubation by placing the particles under running tap water for 3 min, followed by a washing step using approximately 1 L of Milli-Q water. All wet gel particles incubated in DTT and NEM were washed in the same way. After those treatments, the wet gel particles were oven dried and milled as described above (Section 2.2.1). MPs treated with Tgase were made as follows. A 20% w/w Tgase solution was prepared and placed in a fridge at 4 C for 1 week. Before use, the Tgase solution was mixed for 10 min and filtered over a 0.45-mm filter. Tgase was added either to the WPI solution used to make the MPs or to the wet gel particles. In both cases, an enzyme to protein ratio of 1:5 was used. When added to the WPI solution, the Tgase solution was first incubated at 50 C for 10 min and then mixed with the WPI solution. Subsequently, the solution was incubated at 50 C for 1, 2, 6, or 24 h. After incubation, the protein samples were heated and mixed in a bowl mixer, oven dried, and milled as described for the standard MPs (Section 2.2.1). To obtain MPs incubated in a Tgase solution after mixing, 25 g of wet gel particles was added to 50 mL of Tgase solution (preheated for 10 min at 50 C). This dispersion was further mixed at 50 C for
172
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
1, 2, 6, or 24 h, and then oven dried and milled as described in Section 2.2.1. The pH of the WPI solution was lowered before mixing with 1 M HCl to obtain MPs made at pH 5.8. The WPI solution was then heated and mixed, oven dried, and milled as described above for the standard MPs (Section 2.2.1). 2.2.3. Formation of whey protein macro gels WPI macro gels and heated WPI solutions (Section 2.2.4) were made using similar conditions as for the MPs as much as possible. The macro gels were regarded as a model material for the MPs (i.e., the gel was considered as a large MP), because MPs were obtained by drying small gel particles. Drying and rehydrating the macro gels would have mimicked the MPs even better, though then we would not have a large homogenous sample. The macro gels were used to create a calibration curve that related the dry matter content to the transverse relaxation rate R2, so the dry matter content present in each water fraction of the MP pellet could be estimated. To obtain WPI macro gels, solutions of various WPI concentrations were made as described in Section 2.2.1. Then, the solutions were centrifuged to remove air bubbles and carefully transferred to Teflon tubes (diameter 20 mm, length 100 mm). The tubes were heated at 95 C for 50 min, after which they were cooled with running tap water for 5 min. The gels obtained were wrapped in plastic foil and placed in the fridge prior to analysis, which was done within 24 h. WPI macro gels were also made at pH 5.8 and some of the gels were incubated in a DTT solution. The same procedure was used as for the standard WPI macro gels with some small modifications. To obtain gels made at pH 5.8, the pH of the solutions was lowered with 1 M HCl before heating. Gels that had reacted with DTT were made by cutting the gels into pieces of 20 mm in diameter and weighing approximately 2 g. The gels were placed in 25 mL of a 40 mM DTT solution and incubated at 20 C for 24 h. 2.2.4. Formation of unheated and heated whey protein solutions A 10% w/w solution of WPI was made by dissolving WPI in water and mixing the solution at room temperature for 1 h. This solution was then diluted with water to obtain unheated WPI solutions at various concentrations, which were analyzed further. Heated whey protein solutions were made from the unheated WPI solutions by keeping the unheated solutions at 4 C overnight and heating them in a water bath at 95 C for 50 min the next day. Subsequently, the samples were cooled using running tab water for 5 min and analyzed when they were at room temperature. Heated WPI solutions at pH 5.8 were made using a 10% w/w solution prepared as described in Section 2.2.1. The pH of the solution was brought to pH 5.8 with 1 M HCl, after which the solution was further diluted with water to obtain WPI solutions at various concentrations, and centrifuged to remove the air bubbles. After dilution, the pH of the solutions was still 5.8. The solutions were heated at 95 C for 50 min, cooled under running tap water for 5 min and placed in the fridge until they were analyzed within 24 h. 2.2.5. The water-binding capacity of whey protein microparticles The WBC of MPs was determined by preparing 10% dispersions of the MPs. These dispersions were mixed with a vortex until the MPs were hydrated and then mixed with a rotator at a speed of 16 rpm at 25 C for 3 h. Every 15 min, the dispersions were mixed with a vortex again. After hydration, the dispersions were centrifuged at 3000 rpm and 25 C for 20 min to obtain a pellet and supernatant. The supernatant was decanted from the pellet; the weight of the pellet was determined and the WBC of the pellets was calculated using Equation (1). The average dry matter content of the MPs (95%) was used to calculate the dry matter weight of the MPs.
In addition, the MPs were reheated to create pellets with altered WBCs by reheating the dispersions after hydration at 90 C for 30 min. Subsequently, the dispersions were cooled in ice water for 10 min and equilibrated at room temperature for an additional 10 min. Then, the dispersions were centrifuged as described for the unheated MP pellets. The WBC of the pellets used for the calculations on swelling (Section 2.2.7) was measured in triplicate. The WBC of the pellets used for TD NMR analysis (Section 2.2.8.1) was measured once because the results were shown to be reproducible. 2.2.6. Binding of water within the whey microparticles MPs were placed in desiccators for equilibration at various relative humidities (RHs). Saturated salt solutions were made from CH3CO2K, K2CO3, and K2SO4. Approximately 0.2 g of MPs were spread over a plate and placed in the desiccator. The desiccator was set under a vacuum and the samples were kept there for 12 days before they were analyzed. The water activity (aw) of the samples in the desiccator with a CH3CO2K saturated salt solution was 0.25 (25 C), 0.46 (25 C) in a K2CO3 saturated salt solution, and 0.95 (25 C) in a K2SO4 salt solution. This was measured with an Aqualab 4 TE water activity meter (Decagon Devices Inc., Pullman, WA). 2.2.7. Swelling experiments The swelling of MPs in water was determined for hydrated MPs and hydrated and reheated MPs prepared as described in Section 2.2.5. After hydration, or hydration and reheating, the size of the MPs was determined with static light scattering (Mastersizer 2000, Malvern Instruments). Also the size of the unhydrated MPs (i.e., MPs obtained after oven drying and milling and referred to as dry MPs) was determined. Therefore, a 10% w/w dispersion was made with isopropanol, a solvent in which the MPs do not swell. During the measurements, water was used as a carrier fluid for the hydrated MPs and isopropanol for the dry MPs. The rotation speed of the vessel of the carrier fluid was set at 1200 rpm. Refractive indices of 1.545 (MPs), 1.33 (water), and 1.39 (isopropanol) and an absorption of 0.001 (MPs) were used for the calculations €m, Nielsen, Windhab, & Hermansson, 1999). The (Walkenstro measurements were done for three dispersions per type of MP; the size distribution was measured in quintuplicate. The size of the dry MPs was measured in triplicate. From the measurements, the average particle diameter d4,3 (volume mean diameter) was calculated assuming that the MPs were spherical. The average diameter d4,3 was used to calculate the average volume of the dry MPs and the volume increase when hydrated, or hydrated and reheated. The amount of water inside and between the MPs was calculated from the WBC of the pellets, assuming that the proteins did not dissolve and that the density of dry MPs (Purwanti, van der Veen, van der Goot, & Boom, 2013) was 1.3 g cm3 and the density of water was 1 g cm3. 2.2.8. 1H relaxometry 1 H relaxometry was performed with a Maran Ultra NMR spectrometer (at 30.7 MHz proton resonance frequency; Resonance Instruments Ltd., Witney, United Kingdom). T2 relaxation decay curves were recorded by means of a standard Carr-PurcellMeiboom-Gill (CPMG) pulse sequence. The CPMG decay train consisted of 16,384 echoes with an echo time of 1 ms. Experiments were averaged over 16 scans with a repetition time of 30 s. T2 spectra of relative intensity as a function of T2 were determined by numerical inverse Laplace transformation of the data as implemented in CONTIN (Provencher, 1982). The regularization parameter in CONTIN was set to the same, rather conservative value, for all spectra to facilitate comparison of the results. This is justified by the fact that the signal-to-noise ratio was comparable
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
for all measurements. T2 spectra were subsequently analyzed by determining peak positions (the top of the peaks was used to determine the T2 values of the various water fractions in the samples) and areas to determine relative amounts using IDL (ITT Visual Information Solution, Boulder, CO). For decays with a clear mono-exponential character, the LevenbergeMarquardt non-linear least squares algorithm implemented in SPLMOD (Provencher & Vogel, 1983) was used to analyze the data. These decays were measured to obtain T2 values for calibration purposes. The rationale for using SPLMOD instead of CONTIN in those cases is that SPLMOD fits both the amplitude and T2 relaxation time of a discrete sum of exponentials. CONTIN, on the other hand, determines amplitudes given a distribution of fixed T2 values. In the latter, an extra unnecessary step is needed to obtain average T2 values. Every sample was measured once because extra experiments performed on a number of samples showed good reproducibility. This allowed us to measure a broader range of samples. 2.2.8.1. 1H relaxometry of whey protein microparticle pellets. Pellets were made as described in Section 2.2.5. After decanting the supernatant from the pellet, the bottoms of the Eppendorf tubes were cut off so the samples were disturbed as little as possible before analysis (scooping the pellet from the tube affected the results). The tips of the tubes were used for further analysis. The resulting transverse decay curves were all multi-exponential and analyzed with CONTIN. To estimate the absolute amount of water present inside and between the MPs, the amount of water present according to the WBC of the pellets was divided over those two fractions according to the ratio of these areas in the CONTIN T2 spectra. Using the T2 values for the top of the two fractions, the dry matter content inside and between all the various MP pellets was calculated from the standard calibration curves (Supporting information). The calibration curve obtained for solutions and gels made at pH 5.8 (Supporting information) was used only for the MPs made at pH 5.8. From the amount of dry matter that was present between the MPs the percentage of dry matter dissolved from the MPs was calculated. Though, if the mass balance was used the solubility would be lower or larger than calculated via the WBC, T2 values and the calibration curves. The relative area under the peaks was determined by dividing either the area of the peak attributed to water inside the MPs or between the MPs over the sum of the total area.
173
obtained were all assumed to be mono-exponential and analyzed with SPLMOD. 2.2.8.4. 1H relaxometry of unheated and heated whey protein isolate solutions. Unheated and heated WPI solutions, which were made according to Section 2.2.4, were analyzed using TD NMR by pipetting 125 mL of the solution into the NMR tubes. The resulting curves were analyzed with SPLMOD because of the mono-exponential character of the decay curves. 2.2.8.5. 1H relaxometry of whey protein microparticles hydrated in desiccators. After hydration of the MPs in the desiccators (Section 2.2.6), the hydrated MPs were placed in an NMR tube and analyzed. The curves obtained were assumed to be mono-exponential and analyzed with SPLMOD. 2.2.9. Dry matter content The dry matter content of the gels, solutions, and MPs was determined by drying the samples in an oven at 105 C for 24 h. 3. Results and discussion The following sections describe the steps that were taken to gain inside into the water-binding capacity (WBC) of whey protein microparticles (MPs). First, the WBC of the MPs was determined, and then the value was compared to the swelling of the MPs as calculated from the light scattering data. Subsequently, the pellets were analyzed with time domain nuclear magnetic resonance (TD NMR). To be able to further interpret these TD NMR results, swollen MPs were mimicked by MPs subjected to certain relative humidities (RHs) and whey protein gels, and also analyzed with TD NMR. In addition, the dry matter and water concentrations in both water fractions were calculated and compared with the light scattering results. Finally, the water fractions obtained were assigned to the two water domains in the MP pellets (water inside and between the MPs) and the contribution of both water domains to the WBC of the pellet was investigated.
2.2.8.2. 1H relaxometry of supernatants. The spectra of the supernatant of standard MPs, unwashed MPs incubated in 40 mM DTT, and MPs made at pH 5.8 were obtained by making pellets and supernatants first as described in Section 2.2.5. After decanting the supernatant from the pellet, 125 mL was added to the NMR tube and analyzed. The resulting transverse decay curves were assumed to be mono-exponential and analyzed with SPLMOD. The SPLMOD results for the supernatants were used to calculate the dry matter content of the supernatants. For that, only the solution part of the calibration curve was used (Supporting information). For the supernatant of the standard MPs and the unwashed MPs incubated in 40 mM DTT, the calibration curve of the standard solutions was used; for the MPs made at pH 5.8, the curve for the solutions made at pH 5.8 was used. 2.2.8.3. 1H relaxometry of macro gels. After the formation of the macro gels, as described in Section 2.2.3, the gels were cut with a biopsy punch into small cylinders and analyzed. The curves
Fig. 1. The various water domains formed after centrifuging whey protein microparticles (MPs) that were hydrated in an excess amount of water: (A) supernatant, (B) water between the MPs, and (C) water inside the MPs.
174
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
3.1. Analysis of the swelling of whey protein microparticles in a pellet with light scattering To determine the WBC of MPs, the MPs were hydrated in excess water after which the dispersions were centrifuged to obtain a pellet and supernatant (Fig. 1). The pellet, containing swollen MPs and water between the MPs, was weighed to determine the WBC, assuming that the amount of water between the MPs was negligible as an initial hypothesis. However, the importance of water between the MPs is investigated in this study. That is why the term WBC-P (WBC of the pellet) is introduced to distinguish the WBC of the MPs themselves (WBC-MPs) from the WBC of the pellet that includes water between the MPs as well. The WBC-P of the unwashed MPs treated with 40 mM DTT and the MPs made at pH 5.8 differed most from the WBC-P of standard MPs from all tested MPs. Therefore, those three types of MPs were analyzed in more detail. The unwashed MPs treated with 40 mM DTT had a much larger WBC-P than the standard MPs, whereas MPs made at pH 5.8 had a much lower WBC-P (Table 1). However, it is not clear whether this WBC-P can be translated to the WBC-MPs directly or that the water between the MPs contributed to the WBC-P as well. A first route to estimate the relative importance of water inside and between the MPs is to consider the swelling of the MPs. Calculations based on the average particle diameter d4,3 (volume mean diameter) of these MPs (see Section 2.2.7) suggested that the unheated MPs treated with DTT contained more water than the standard MPs, whereas both the unheated and reheated MPs made at pH 5.8 contained less water than the standard MPs (Table 1). Although these results are indicative (among others, because the particles were not spherical (Peters et al., 2015)), they strongly suggest that swelling of the MPs only cannot explain the WBC-P. In all cases, the water content inside the swollen MPs was significantly lower than the total amount of water present in the pellets (WBC-P). This implies that a certain amount of water was present between the MPs since the pellet consists of water inside and between the MPs. The WBC-MPs is therefore most probably lower than the WBC-P.
Fig. 2. The CONTIN T2 spectra of (A) pellets and (B) supernatant formed during the water-binding experiment with standard whey protein microparticles (ST MPs) (solid line), unwashed MPs treated with 40 mM DTT (DTT MPs) (dashed line), and MPs made at pH 5.8 (pH MPs) (dotted line).
fraction appearing around T2 ¼ 1 s seems to be related to water that is not bound within the pellet. The T2 value of both the supernatant and expelled water fraction was lower than the T2 value of free water, which was measured to be around 2.4 to 2.7 s, suggesting the presence of proteins in the supernatant and expelled water. The other two fractions are therefore ascribed to water present inside the pellets. The T2 values of the fractions with the smallest transverse relaxation time (T2,1) were different for the three samples. A smaller WBC corresponded to a smaller T2,1, suggesting a larger dry matter content in that water fraction. The second fraction (T2,2) had a T2 value around 0.1 s, which was similar for all three pellets. The ratio of the areas with T2,1 and T2,2 differed between the pellets as depicted in Fig. 2A. The unwashed MPs treated with 40 mM DTT had a larger area for the T2,2 fraction than for the T2,1
3.2. Analysis of pellets of whey protein microparticles with time domain nuclear magnetic resonance Time domain nuclear magnetic resonance (TD NMR) was used to further analyze the water distribution in the pellets. CONTIN transverse relaxation time (T2) spectra for the pellets of standard MPs, unwashed MPs treated with 40 mM DTT, and MPs made at pH 5.8 showed two main peaks and one small peak around T2 ¼ 1 s (Fig. 2A), indicating the presence of three different water fractions. The fraction with the largest transverse relaxation time (T2,3) was assigned to expelled water that was present after cutting the tube with the pellet. This assignment was done based on the fact that the supernatant obtained after centrifugation (domain A in Fig. 1) also showed a T2 fraction around 1 s (Fig. 2B). Therefore, the water
Table 1 The water-binding capacity (WBC) of the unheated and reheated pellets (g water/g dry matter) of standard (ST) whey protein microparticles (MPs), unwashed MPs treated with 40 mM dithiothreitol (DTT MPs) and MPs made at pH 5.8 (pH MPs), the average size of the dry, hydrated, and hydrated and reheated MPs (mm) (d4,3), and the calculated water content (g water/g dry matter) and dry matter content (%) inside the unheated and reheated swollen MPs. WBC pellets (g water/g dry matter)
Average size MPs (mm) (d4,3)
Unheated
Reheated
Dry
5.4 (0.1) 7.9 (0.0) 2.4 (0.1)
79 (2) 110 (2) 78 (2) 120 (3) 74 (3) 99 (4)
ST MPs 3.7 (0.2) DTT MPs 6.9 (0.2) pH MPs 2.1 (0.2)
Calculated water content inside swollen MPs (g Calculated dry matter content of water/g dry matter) swollen MPs (%)
Hydrated Hydrated and reheated Unheated
The standard deviations are given in parentheses.
127 (2) 114 (3) 98 (2)
1.3 2.0 1.1
Reheated
Unheated
Reheated
2.4 1.6 1.0
43 33 48
29 38 50
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
175
fraction; the areas for these fractions were nearly the same for the MPs made at pH 5.8. For further insight into the WBC-P, the pellets of all the other MPs (MPs treated with DTT washed and unwashed, MPs treated with DTT and N-ethylmaleimide, MPs treated with genipin, and MPs treated with transglutaminase (Tgase) by addition of Tgase to the WPI solution or the wet gel particles), both unheated and reheated, were analyzed with TD NMR. The CONTIN T2 spectra of all MP pellets showed at least the two water fractions T2,1 and T2,2. Fig. 3A shows that T2,1 increased from WBC-P 2 to 5.5 g water/g dry matter, after which it tends to deviate from the trend. No clear relationship was found between T2,2 and the WBC-P (Fig. 3B), because all pellets had a T2,2 value around 0.12 (±0.02). The relative area2,2 was almost linearly related to the WBC-P (Fig. 3C). Therefore, these results suggest that an increase in WBC-P is caused by both an increase in the amount of water in water fraction T2,1 (T2,1 increased with an increase in the WBC implying an increase in the water content) and in water fraction T2,2 (the relative area2,2 increased with an increase in WBC). 3.3. Assignment of water fractions obtained with time domain nuclear magnetic resonance to the water domains in whey protein microparticle pellets It is likely that T2,1 and T2,2 can be assigned to the two water domains within the pellet: the swollen MPs and water between the MPs. The exchange rate of protons and the water/protein ratio are probably larger inside the MPs (domain C in Fig. 1) than between the MPs (domain B in Fig. 1). Therefore, it can be hypothesized that T2,1 represents water inside the MPs and T2,2 the water between the MPs. However, from the TD NMR results of the pellets only it is not clear if this assignment can be done instantly, or that water in the pellet was divided over two water peaks for other reasons. Therefore, it would be the best to analyze swollen MPs and water between the MPs separately. Since it would be hard to separate both fractions from the pellet, here it was tried to mimic swollen MPs to get to know if T2,1 could be assigned to this water domain. To get hydrated MPs without water between the MPs, MPs were placed in a desiccator with controlled RHs for 12 days. A TD NMR analysis of those MPs resulted in spectra with mainly one fraction, as illustrated in Fig. 4A for the standard MPs. Smaller T2 values were found when less water was present. The T2 value of the MPs at RH 95% was still one order smaller than the T2,1 value of the pellet spectra. This relatively large difference corresponds to the isotherm of these MPs, which shows a strong increase in moisture content from aw 0.95 to 1 (Peters et al., 2015). In addition, whey protein isolate (WPI) macro gels were analyzed with TD NMR, because macro gels were considered as a model material for MPs without surrounding water but with a large moisture content. Fig. 4B shows that all the spectra for WPI macro gels contained one main fraction in the same range as the T2,1 fraction of the pellets (Fig. 2A). Mono-exponential behavior for WPI and whey protein concentrate gels with a protein concentration of 6 to 36% (w/w) has also been found by others (Colsenet, Mariette, & Cambert, 2005; Lambelet, Berrocal, Desarzens, Froehlicher, & Ducret, 1988; Oztop, Rosenberg, Rosenberg, McCarthy, McCarty, 2010). As expected, the T2 value of the gels decreased when the WPI concentration of the gel increased (Fig. 4B), because an increase in the protein content increases the total interaction with water and reduces the transverse relaxation time (Oakes, 1976b). The spectra of the WPI macro gels and MPs hydrated in a desiccator suggest both that one fraction (T2,1) within the pellet spectra is associated with water inside the MPs, and that therefore T2,2 can be assigned to water between the MPs.
Fig. 3. Relationship between the water-binding capacity (WBC-P) (g water/g dry matter) of the unheated (filled diamond) and reheated (open diamond) pellets and (A) T2,1 (s), (B) T2.2 (s), and (C) the relative area2,2 (-).
Another method to check whether T2,1 can be assigned to water inside the MPs and T2,2 to water between the MPs is by calculating the water and dry matter concentrations within those domains and comparing those concentrations to the light scattering data. Therefore, a standard calibration curve relating the dry matter content to the transverse relaxation rate R2 (R2 ¼ 1/T2) was drawn (Supporting information) (Colsenet et al., 2005; Lambelet et al., 1988; Oakes, 1976a, 1976b). A separate calibration curve was made for the MPs made at pH 5.8, because those samples showed a different relationship between R2 and the dry matter content (Supporting information). The dry matter content in the supernatant of the standard MPs and MPs made at pH 5.8 could be estimated well (Table 2), indicating the validity of the calibration
176
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
Fig. 5. Relationship between the water-binding capacity of the pellets (WBC-P) (g water/g dry matter) and the amount of water (g water/g dry matter) present inside (filled diamond) and between (open diamond) the whey protein microparticles (MPs).
Fig. 4. The CONTIN T2 spectra of (A) the standard microparticles incubated in a desiccator with RH 25% (dotted line), 46% (dashed line), and 95% (solid line); and (B) whey protein isolate (WPI) gels with a WPI concentration (w/w) of 19% (solid line), 30% (dashed line), and 39% (dotted line).
curves. The supernatant values obtained for the MPs treated with 40 mM DTT show some discrepancy between the measured and calculated dry matter content. These samples seem to be a special case as they also deviate from the trend in Fig. 3A. The dry matter content inside the MPs estimated from the TD NMR data and the calibration curves (Table 2) compares well with the light scattering data for the standard MPs, the unwashed MPs incubated in 40 mM DTT and the MPs made at pH 5.8 (Table 1). So, it was concluded that T2,1 can be assigned to water inside the MPs and T2,2 to water between the MPs. 3.4. Contribution of water inside and between the whey protein microparticles to the water-binding capacity of the pellets The next step is to estimate the contribution of water inside and between the MPs to the WBC-P. Therefore, the total amount of water present in the pellets is divided into two water fractions according to the ratio between the areas under the curve of those two fractions. This division shows that the amount of water present inside the MPs increased slightly in the WBC-P ranging from 2 to 5.5 g water/g dry matter (Fig. 5). It seems that the increase in T2,1 with increasing WBC-P was less important than the decrease in the relative area2,1 with increasing WBC-P. The increase in the WBC-P appears to be mainly related to an increase in the amount of water between the MPs. A consequence of that reasoning is that the contribution of water between the MPs cannot be neglected in the WBC-P, and that the WBC-MPs cannot be determined via the WBC-P.
The contribution of water between the MPs to the WBC-P seems to be large. At a WBC of 5.5 g water/g dry matter 80% of the water was present between the MPs, which implies that water-binding in this fraction in the pellet probably consisted of more than water inclusion by the MPs only. A remarkable observation for all MP pellets is that they had nearly the same T2,2 (Fig. 3B), which suggests a similar dry matter content between the MPs. Consequently, it seems that dissolved proteins originating from the MPs are a determining factor for the WBC-P. If it is assumed that the T2 of water is caused by dissolved proteins only, the T2,2 values suggest the presence of around 11 to 15% dry matter between the MPs. Although a certain amount of dry matter was expected due to the fact that the supernatant contains proteins as well (Fig. 2B and Table 2), this high dry matter content was unexpected. The suggested higher dry matter content between the MPs could be related to small highly swollen protein aggregates that were dragged down on centrifugation. The protein content suggested by the T2,2 value could also explain the binding of water between the MPs, because this concentration might have led to the formation of a (particle) gel, or an increased viscosity at least, upon centrifugation. Without such effect, more water would be expelled during centrifugation. Rough calculations show that the volume fraction of standard MPs in a pellet is around 0.5 only, which is clearly lower than the maximum packing density of randomly packed monodisperse spheres (around 0.65) while polydisperse systems could have an even higher packing density (Walstra, 2003). Nevertheless, the possibility that T2,2 was influenced by other interactions and/or capillary pressures cannot be excluded. In that case, the actual protein concentration is somewhat lower than indicated by T2,2, but still much larger than the protein content in the supernatant. Fig. 5 also shows the different behavior of the unwashed DTT treated particles (the particles with a WBC-P larger than 5.5 g water/g protein), compared to the other MPs. The differences are probably due to the fact that no clear distinction can be made between water inside and between those particular MPs due to overlapping water peaks in the TD NMR spectra. This seems likely, because the area of T2,1 for the unwashed MPs incubated in 40 mM DTT was small compared with the area of T2,1 for the other MPs (Fig. 2A). It may be that those MPs incubated in 40 mM DTT partly solubilized, leading to a nearly equal dry matter and water content inside and between the MPs. After the division of water according to the ratio of the areas, the WBC-P was calculated from the concentrations obtained via calibration curves. A comparison between the WBC-P calculated from
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
177
Table 2 The measured dry matter content (%) (standard deviation) and calculated dry matter content (%) with the calibration curves in the supernatants of the unheated and reheated standard (ST) whey protein microparticles (MPs), unwashed MPs treated with 40 mM DTT (DTT MPs), and MPs made at pH 5.8 (pH MPs), and the calculated dry matter content (%) and solubility of these MPs (%) in the unheated and reheated pellets. Supernatant
ST MPs DTT MPs pH MPs
Pellet
Measured dry matter content (%)
Calculated dry matter content (%)
Calculated dry matter content (%)
Unheated
Unheated
Unheated
0.6 (0.1) 2.2 (0.5) 0.5 (0.1)
Reheated
0.6 (0.1) 3.2 (0.5) 0.5 (0.1)
0.7 1.2 0.4
Reheated
0.9 2.1 0.4
the TD NMR data and the measured WBC-P shows that the WBC-P measured from the TD NMR data was mostly larger than the measured WBC-P (Fig. 6). This is probably caused by the way the WBC-P was determined when using the pellet weight after centrifugation. This method does not consider the solubilization of dry matter, and therefore it can overestimate the dry matter content and consequently underestimate the WBC. A 10% smaller dry matter content can explain the difference between the measured WBC-P and the calculated WBC-P from the TD NMR data as shown in Fig. 6 with the dotted line. With TD NMR, on the other hand, the pellet obtained after centrifugation was used for analysis and the water concentration was obtained from measuring 1H protons. This has probably resulted in more accurate values for the dry matter and water concentrations within both water domains in the pellets and a more accurate WBC-P. Therefore, it can be concluded that with TD NMR a distinction between water present inside and between the MPs in a pellet can be made. In addition, TD NMR is a more exact method to determine the WBC-P and the dry matter and water concentrations in the water fractions of the pellet than the use of the pellet weight obtained after centrifugation of a MP dispersion. 4. Conclusions The water-binding capacity (WBC) of protein samples is frequently determined from the pellet weight obtained after centrifuging a dispersion of this sample, assuming that water
Solubility of MPs (%)
Reheated
Between MPs
Inside MPs
Between MPs
Inside MPs
12 12 12
35 31 59
9 12 8
25 27 60
Unheated
Reheated
33 89 23
39 87 15
between the sample particles can be neglected. However, we have demonstrated in this paper that the amount of water between the particles cannot be neglected. For pellets made of whey protein microparticle (MP) dispersions, with time domain nuclear magnetic resonance (TD NMR) a distinction was made between water inside and between the MPs. It was shown that an increase in WBC of the pellet (WBC-P) from 2 to 5.5 g water/g dry matter was partly caused by a slight increase in swelling of the MPs, but to a larger extent by an increase in the amount of water between the MPs. The latter could account for as much as 80% of the water being kept in a pellet with a WBC of 5.5 g water/g dry matter. Therefore, it is concluded that the WBC-P is not a measure for the WBC of MPs (WBC-MPs); the outcomes of the centrifugation method should be interpreted cautiously when used to determine the WBC-MPs. We hypothesize that this relatively large amount of water between the MPs could be present because small highly swollen protein structures co-sedimented between the MPs during centrifugation. The presence of such structures can also explain the remarkable result of a nearly constant T2 value at all WBC-Ps, which suggest the presence of 11 to 15% of dry matter in the water fraction of water between the MPs. The results presented in this paper demonstrate that TD NMR is a better method to determine the WBC-MP, the amount of water between the MPs and the overall WBC-P than using the pellet weight after centrifugation. We therefore conclude that TD NMR used in the manner proposed is a useful additional tool to understand WBC of a dispersed particle system. Acknowledgments The authors would like to thank FrieslandCampina and NanoNextNL, a consortium of the government of the Netherlands and 130 partners, for their financial support of this research. In addition, we would like to thank Antoinette Toebes of FrieslandCampina for help with the TD NMR experiments and reading of the manuscript. Appendix. Supporting information Supporting information related to this article can be found at http://dx.doi.org/10.1016/j.foodhyd.2015.09.031. References
Fig. 6. The measured water-binding capacity of the pellets (WBC-P) (g water/g dry matter) and the WBC-P calculated from time domain nuclear magnetic resonance (TD NMR data) of the unheated (filled diamond) and reheated (open diamond) MPs. The WBC-P was calculated from the ratio between the areas of water inside (T2,1) and outside (T2,2) the MPs from the CONTIN T2 spectra and the protein concentrations obtained from the calibrations curves. The solid line shows the relation if the measured WBC-P was the same as the TD NMR WBC-P. The dashed line shows the relation in case the dry matter content of the pellets was actually 10% higher than used to calculate the measured WBC-P.
Ahmed, J., Al-Jassar, S., & Thomas, L. A. (2015). Comparison in rheological, thermal, and structural properties between Indian Basmati and Egyptian Giza rice flour dispersions as influenced by particle size. Food Hydrocolloids, 48, 72e83. Berghout, J. A. M., Boom, R. M., & van der Goot, A. J. (2014). The potential of aqueous fraction of lupin seeds for high-protein foods. Food Chemistry, 159, 64e70. Bertram, H. C., Andersen, H. J., & Karlsson, A. H. (2001). Comparative study of lowfield NMR relaxation measurements and two traditional methods in the determination of water holding capacity of pork. Meat Science, 57(2), 125e132. Bertram, H. C., Dønstrup, S., Karlsson, A. H., & Andersen, H. J. (2002). Continuous distribution analysis of T2 relaxation in meatdan approach in the determination of water-holding capacity. Meat Science, 60(3), 279e285.
178
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
Brøndum, J., Munck, L., Henckel, P., Karlsson, A., Tornberg, E., & Engelsen, S. B. (2000). Prediction of water-holding capacity and composition of porcine meat by comparative spectroscopy. Meat Science, 55(2), 177e185. Colsenet, R., Mariette, F., & Cambert, M. (2005). NMR relaxation and water selfdiffusion studies in whey protein solutions and gels. Journal of Agricultural and Food Chemistry, 53(17), 6784e6790. van Duynhoven, J., Voda, A., Witek, M., & Van As, H. (2010). Time-domain NMR applied to food products. In A. W. Graham (Ed.), Annual reports on NMR spectroscopy (Vol. 69, pp. 145e197). Academic Press. Fennema, O. R. (1996). Food chemistry (3rd ed., pp. 366e370). Taylor & Francis. Goetz, J., & Koehler, P. (2005). Study of the thermal denaturation of selected proteins of whey and egg by low resolution NMR. LWT e Food Science and Technology, 38(5), 501e512. € tz, J., Noll, M., Wolfschoon, A., Eibel, H., & Weisser, H. (2004). Hinrichs, R., Go Characterisation of the water-holding capacity of fresh cheese samples by means of low resolution nuclear magnetic resonance. Food Research International, 37(7), 667e676. €tz, J., & Weisser, H. (2003). Water-holding capacity and structure of Hinrichs, R., Go hydrocolloid-gels, WPC-gels and yogurts characterised by means of NMR. Food Chemistry, 82(1), 155e160. Jain, A., Prakash, M., & Radha, C. (2015). Extraction and evaluation of functional properties of groundnut protein concentrate. Journal of Food Science and Technology. http://dx.doi.org/10.1007/s13197-015-1758-7. Kneifel, W., Paquin, P., Abert, T., & Richard, J. P. (1991). Water-holding capacity of proteins with special regard to milk proteins and methodological aspectsda review. Journal of Dairy Science, 74(7), 2027e2041. Lambelet, P., Berrocal, R., Desarzens, C., Froehlicher, I., & Ducret, F. (1988). Pulsed low-resolution NMR investigations of protein sols and gels. Journal of Food Science, 53(3), 943e946. Mariette, F. (2006). NMR relaxation of dairy products. In G. A. Webb (Ed.), Modern magnetic resonance (pp. 1697e1701). Springer. Mitchell, J., Gladden, L. F., Chandrasekera, T. C., & Fordham, E. J. (2014). Low-field permanent magnets for industrial process and quality control. Progress in Nuclear Magnetic Resonance Spectroscopy, 76, 1e60. Oakes, J. (1976a). Protein hydration. Nuclear magnetic resonance relaxation studies of the state of water in native bovine serum albumin solutions. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
72, 216e227. Oakes, J. (1976b). Thermally denatured proteins. Nuclear magnetic resonance, binding isotherm and chemical modification studies of thermally denatured bovine serum albumin. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 72, 228e237. Oztop, M. H., Rosenberg, M., Rosenberg, Y., McCarthy, K. L., & McCarthy, M. J. (2010). Magnetic resonance imaging (MRI) and relaxation spectrum analysis as methods to investigate swelling in whey protein gels. Journal of Food Science, 75(8), E508eE515. Peters, J. P. C. M., Luyten, H., Alting, A. C., Boom, R. M., & van der Goot, A. J. (2015). Effect of crosslink density on the water-binding capacity of whey protein microparticles. Food Hydrocolloids, 44, 277e284. Provencher, S. W. (1982). A constrained regularization method for inverting data represented by linear algebraic or integral equations. Computer Physics Communications, 27(3), 213e227. Provencher, S. W., & Vogel, R. H. (1983). Regularization techniques for inverse problems in molecular biology. In P. Deuflhard, & E. Hairer (Eds.), Numerical treatment of inverse problems in differential and integral equations (Vol 2, pp. 304e319). Boston: Birkh€ auser. Purwanti, N., van der Veen, E., van der Goot, A. J., & Boom, R. (2013). Stiffening in gels containing whey protein isolate. International Dairy Journal, 28(2), 62e69. Straadt, I. K., Rasmussen, M., Andersen, H. J., & Bertram, H. C. (2007). Aging-induced changes in microstructure and water distribution in fresh and cooked pork in relation to water-holding capacity and cooking loss e a combined confocal laser scanning microscopy (CLSM) and low-field nuclear magnetic resonance relaxation study. Meat Science, 75(4), 687e695. Sun, Q., & Xiong, C. S. L. (2014). Functional and pasting properties of pea starch and peanut protein isolate blends. Carbohydrate Polymer, 101, 1134e1139. €m, P., Nielsen, M., Windhab, E., & Hermansson, A. M. (1999). Effects of Walkenstro flow behaviour on the aggregation of whey protein suspensions, pure or mixed with xanthan. Journal of Food Engineering, 42(1), 15e26. Walstra, P. (2003). Physical chemistry of foods (p. 99). Marcel Dekker, Inc. Yu, J., Ahmedna, M., & Goktepe, I. (2007). Peanut protein concentrate: production and functional properties as affected by processing. Food Chemistry, 103(1), 121e129. Zayas, J. F. (1997). Functionality of proteins in food (pp. 76e133). Springer Berlin Heidelberg.
Contents lists available at ScienceDirect
Food Hydrocolloids journal homepage: www.elsevier.com/locate/foodhyd
Time domain nuclear magnetic resonance as a method to determine and characterize the water-binding capacity of whey protein microparticles Jorien P.C.M. Peters a, Frank J. Vergeldt b, Henk Van As b, Hannemieke Luyten c, Remko M. Boom b, Atze Jan van der Goot a, * a b c
Food Process Engineering Group, Wageningen University, PO Box 18, 6700 AA Wageningen, The Netherlands Laboratory of Biophysics and Wageningen NMR Centre, Wageningen University, Dreijenlaan 3, 6703 HA Wageningen, The Netherlands FrieslandCampina, Stationsplein 4, 3818 LE Amersfoort, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 July 2015 Received in revised form 24 September 2015 Accepted 30 September 2015 Available online 9 October 2015
Water-binding capacity (WBC) is commonly measured with a centrifugation method in which a sample is hydrated in excess water and the pellet weight after centrifugation defines the WBC. When a dispersion is being analyzed, here containing whey protein microparticles (MPs), the pellet consists of swollen particles and water between the particles. These two water domains in MP pellets were distinguished using time domain nuclear magnetic resonance (TD NMR). This distinction showed that an increase in WBC from 2 to 5.5 g water/g dry matter was mainly due to an increase in water between the MPs. Besides, it was found that TD NMR-measurements could be used to provide accurate values of the amount of water in both water domains in MP pellets. This makes TD NMR therefore a more accurate method to determine the WBC of the whole pellet than weighing the pellet after centrifugation. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Proteins Water-binding capacity Swelling Nuclear magnetic resonance Centrifugation Microstructure
1. Introduction An important functional property of proteins is their ability to bind water. Unfortunately, this property is not well defined; various definitions and terms are used in the literature (Fennema, 1996; Kneifel, Paquin, Abert, & Richard, 1991; Zayas, 1997). Here, we use the term water-binding capacity (WBC) to describe the ability of a protein sample present in excess water to bind water when subjected to an external force. Several methods can be used to determine the WBC of a protein sample, but commonly a centrifugation method is used in which the obtained pellet weight determines the WBC. In this method, the protein sample is first placed in excess water to hydrate, and then surplus water is separated from the proteins by centrifugation. The WBC is then calculated using the
* Corresponding author. E-mail address: [email protected] (A.J. van der Goot). http://dx.doi.org/10.1016/j.foodhyd.2015.09.031 0268-005X/© 2015 Elsevier Ltd. All rights reserved.
following equation:
WBC ðg water=g dry matterÞ ¼
weightpellet weightdry matter weightdry matter (1)
where weightpellet is the pellet weight after centrifugation and weightdry_matter is the weight of the dry matter of the particles. In this study, water-binding of whey protein microparticles (MPs) is investigated to better understand the contribution of swelling of MPs and the water between MPs on the WBC. Previous research with MPs (Peters, Luyten, Alting, Boom, & van der Goot, 2015) suggested that the amount of water bound by the pellet is not necessarily a measure of the amount of water bound by the MPs, because a certain amount of water is bound between the MPs. Therefore, more insight is required into the ratio of water inside and between MPs. Because the pellet weight obtained after
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
centrifugation is used for other dispersions as well to determine the WBC (Ahmed, Al-Jassar, & Thomas, 2015; Berghout, Boom, & van der Goot, 2014; Jain, Prakash, & Radha, 2015; Sun & Xiong, 2014; Yu, Ahmedna, & Goktepe, 2007), the results of this study will have broader applicability and consequences. Time domain nuclear magnetic resonance (TD NMR) or relaxometry seemed to be a promising method to gain more understanding of WBC for MP pellets (van Duynhoven, Voda, Witek, & €tz, & Weisser, Van As, 2010; Goetz & Koehler, 2005; Hinrichs, Go 2003; Hinrichs et al., 2004; Mitchell, Gladden, Chandrasekera, & Fordham, 2014). TD NMR is a non-destructive method and is used to determine, inter alia, the transverse relaxation time (T2) of water via the transverse relaxation of 1H protons (Mitchell et al., 2014). T2 or the transverse relaxation rate R2 (R2 ¼ 1/T2) of every water fraction provides insight into the degree of exchange of either water protons with protein protons or the exchange of water between water fractions such as bulk water with the immobile water fraction around the proteins (Mariette, 2006). Since several mechanisms, such as diffusive and chemical exchange, and crossrelaxation, affect the T2 at the same time, a direct interpretation of the TD NMR results is difficult (Mariette, 2006). However, these mechanisms also cause that the T2 of water in a protein sample is related to the local protein and water concentrations, and the nanoand microstructure (e.g. pores and capillaries) of the protein network (Mariette, 2006; Oakes, 1976a, 1976b). Because both the protein concentration and the protein network structure are different inside and between the MPs in a pellet (Peters et al., 2015), we hypothesize that TD NMR can be used to distinguish between these two water domains. Several studies on meat (Bertram, Andersen, & Karlsson, 2001; Bertram, Dønstrup, Karlsson, & Andersen, 2002; Brøndum et al., 2000; Straadt, Rasmussen, Andersen, & Bertram, 2007) investigated the applicability of TD NMR to measure and gain insight into the WBC and WHC (e.g. the ability of meat to hold water under gravitational force). A good correlation was found between T2 of the water fraction assigned to the weakest bound water in meat and its WBC or WHC. In addition, the amplitude of this water fraction had a good correlation with the WBC or WHC. Therefore, it was concluded that TD NMR is a suitable method to gain insight into the WBC and WHC of meat. In addition to meat, TD NMR was found to be useful to examine the WHC of fresh cheeses (Hinrichs et al., 2004), carrageenan gels and solutions, whey gels, and yoghurt (Hinrichs et al., 2003). Though, to our knowledge, pellets obtained after the frequently used method of centrifuging protein dispersions to measure the WBC of those proteins have not been investigated yet. In this study, several modified MPs with different WBCs were made. The WBC of these MPs was determined through centrifugation of a dispersion of MPs and determining the pellet weight. The MPs and the pellet formed after centrifugation were further characterized with light scattering and TD NMR. These results were related to the WBC of the pellets to better understand the relationship between the swelling capacity of the MPs, the amount of water between the MPs, and the WBC of the pellets of MPs. In addition, a method is shown how TD NMR can be used to determine the WBC of MP pellets. 2. Materials and methods 2.1. Materials Whey protein isolate (WPI; BiPRO, lot no. JE 034-7-440, Davisco Food International Inc., Le Sueur, MN) was used to produce whey protein microparticles (MPs), gels, and solutions. WPI has a reported protein content of 97.6% on a dry basis. The water-binding
171
capacity (WBC) of the MPs was changed by treatment with dithiothreitol (DTT), N-ethylmaleimide (NEM), genipin (all from SigmaeAldrich, Germany), HCl, or microbial Ca2þ-independent transglutaminase (Tgase; Activa WM; activity of 81e135 units g1 according to the manufacturer) derived from Streptoverticillium mobaraense (Barentz Raw Materials, the Netherlands). Isopropanol Emplura (Merck, Germany) was used as a carrier fluid in static light scattering measurements. Potassium acetate (CH3CO2K), potassium carbonate (K2CO3), and potassium sulfate (K2SO4) (synthesis grade; SigmaeAldrich, Germany) were used to make saturated salt solutions. Milli-Q water was used (resistivity of 18.2 MU cm at 25 C, total oxidizable carbon <10 ppb; Merck Millipore, France) in all experiments, unless stated otherwise. 2.2. Methods 2.2.1. Formation of standard whey protein microparticles Standard MPs were made as described in Peters et al. (2015). WPI was dissolved in water to obtain a 40% w/w solution. This solution was mixed using an overhead mixer at 100 rpm and room temperature for at least 2 h, and then mixed further with a magnetic stirrer at 4 C overnight. The solution was centrifuged at 1000 rpm and 20 C for 10 min. After removing the foam layer manually, the solution was mixed in a bowl mixer (type W50) connected to a Brabender Do-corder E330 (Brabender OHG, Duisburg, Germany). The mixer was heated with water at 95 C using a water bath. The mixer settings were as follows: 5 min at 0 rpm, 5 min at 5 rpm, and 40 min at 200 rpm. After mixing, the mixer was cooled with water from a 4 C water bath for approximately 5 min. The wet gel particles were taken out of the mixer and placed in an oven at 50 C for 2 days. The dried gels were milled with an ultracentrifuge mill (Retsch ZM 1000) equipped with a 24-tooth stainless steel rotor. An 80-mm sieve was attached to the rotor, which operated at 15,000 rpm for 120 s. The particles obtained after milling were the standard MPs. 2.2.2. Formation of whey protein microparticles with altered waterbinding capacities MPs with an altered WBC were made by adapting the standard procedure (Section 2.2.1). MPs treated with DTT, a combination of DTT and NEM, or genipin were made by adding 25 g of wet gel particles to a 50-mL solution containing the chemicals. Concentrations of 10, 20, or 40 mM DTT were used with or without 15 mM NEM, and 0.5, 1, 2, or 4 mM of genipin. The dispersions were incubated and mixed at 20 C for 24 h. Part of the wet gel particles with DTT were washed after incubation by placing the particles under running tap water for 3 min, followed by a washing step using approximately 1 L of Milli-Q water. All wet gel particles incubated in DTT and NEM were washed in the same way. After those treatments, the wet gel particles were oven dried and milled as described above (Section 2.2.1). MPs treated with Tgase were made as follows. A 20% w/w Tgase solution was prepared and placed in a fridge at 4 C for 1 week. Before use, the Tgase solution was mixed for 10 min and filtered over a 0.45-mm filter. Tgase was added either to the WPI solution used to make the MPs or to the wet gel particles. In both cases, an enzyme to protein ratio of 1:5 was used. When added to the WPI solution, the Tgase solution was first incubated at 50 C for 10 min and then mixed with the WPI solution. Subsequently, the solution was incubated at 50 C for 1, 2, 6, or 24 h. After incubation, the protein samples were heated and mixed in a bowl mixer, oven dried, and milled as described for the standard MPs (Section 2.2.1). To obtain MPs incubated in a Tgase solution after mixing, 25 g of wet gel particles was added to 50 mL of Tgase solution (preheated for 10 min at 50 C). This dispersion was further mixed at 50 C for
172
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
1, 2, 6, or 24 h, and then oven dried and milled as described in Section 2.2.1. The pH of the WPI solution was lowered before mixing with 1 M HCl to obtain MPs made at pH 5.8. The WPI solution was then heated and mixed, oven dried, and milled as described above for the standard MPs (Section 2.2.1). 2.2.3. Formation of whey protein macro gels WPI macro gels and heated WPI solutions (Section 2.2.4) were made using similar conditions as for the MPs as much as possible. The macro gels were regarded as a model material for the MPs (i.e., the gel was considered as a large MP), because MPs were obtained by drying small gel particles. Drying and rehydrating the macro gels would have mimicked the MPs even better, though then we would not have a large homogenous sample. The macro gels were used to create a calibration curve that related the dry matter content to the transverse relaxation rate R2, so the dry matter content present in each water fraction of the MP pellet could be estimated. To obtain WPI macro gels, solutions of various WPI concentrations were made as described in Section 2.2.1. Then, the solutions were centrifuged to remove air bubbles and carefully transferred to Teflon tubes (diameter 20 mm, length 100 mm). The tubes were heated at 95 C for 50 min, after which they were cooled with running tap water for 5 min. The gels obtained were wrapped in plastic foil and placed in the fridge prior to analysis, which was done within 24 h. WPI macro gels were also made at pH 5.8 and some of the gels were incubated in a DTT solution. The same procedure was used as for the standard WPI macro gels with some small modifications. To obtain gels made at pH 5.8, the pH of the solutions was lowered with 1 M HCl before heating. Gels that had reacted with DTT were made by cutting the gels into pieces of 20 mm in diameter and weighing approximately 2 g. The gels were placed in 25 mL of a 40 mM DTT solution and incubated at 20 C for 24 h. 2.2.4. Formation of unheated and heated whey protein solutions A 10% w/w solution of WPI was made by dissolving WPI in water and mixing the solution at room temperature for 1 h. This solution was then diluted with water to obtain unheated WPI solutions at various concentrations, which were analyzed further. Heated whey protein solutions were made from the unheated WPI solutions by keeping the unheated solutions at 4 C overnight and heating them in a water bath at 95 C for 50 min the next day. Subsequently, the samples were cooled using running tab water for 5 min and analyzed when they were at room temperature. Heated WPI solutions at pH 5.8 were made using a 10% w/w solution prepared as described in Section 2.2.1. The pH of the solution was brought to pH 5.8 with 1 M HCl, after which the solution was further diluted with water to obtain WPI solutions at various concentrations, and centrifuged to remove the air bubbles. After dilution, the pH of the solutions was still 5.8. The solutions were heated at 95 C for 50 min, cooled under running tap water for 5 min and placed in the fridge until they were analyzed within 24 h. 2.2.5. The water-binding capacity of whey protein microparticles The WBC of MPs was determined by preparing 10% dispersions of the MPs. These dispersions were mixed with a vortex until the MPs were hydrated and then mixed with a rotator at a speed of 16 rpm at 25 C for 3 h. Every 15 min, the dispersions were mixed with a vortex again. After hydration, the dispersions were centrifuged at 3000 rpm and 25 C for 20 min to obtain a pellet and supernatant. The supernatant was decanted from the pellet; the weight of the pellet was determined and the WBC of the pellets was calculated using Equation (1). The average dry matter content of the MPs (95%) was used to calculate the dry matter weight of the MPs.
In addition, the MPs were reheated to create pellets with altered WBCs by reheating the dispersions after hydration at 90 C for 30 min. Subsequently, the dispersions were cooled in ice water for 10 min and equilibrated at room temperature for an additional 10 min. Then, the dispersions were centrifuged as described for the unheated MP pellets. The WBC of the pellets used for the calculations on swelling (Section 2.2.7) was measured in triplicate. The WBC of the pellets used for TD NMR analysis (Section 2.2.8.1) was measured once because the results were shown to be reproducible. 2.2.6. Binding of water within the whey microparticles MPs were placed in desiccators for equilibration at various relative humidities (RHs). Saturated salt solutions were made from CH3CO2K, K2CO3, and K2SO4. Approximately 0.2 g of MPs were spread over a plate and placed in the desiccator. The desiccator was set under a vacuum and the samples were kept there for 12 days before they were analyzed. The water activity (aw) of the samples in the desiccator with a CH3CO2K saturated salt solution was 0.25 (25 C), 0.46 (25 C) in a K2CO3 saturated salt solution, and 0.95 (25 C) in a K2SO4 salt solution. This was measured with an Aqualab 4 TE water activity meter (Decagon Devices Inc., Pullman, WA). 2.2.7. Swelling experiments The swelling of MPs in water was determined for hydrated MPs and hydrated and reheated MPs prepared as described in Section 2.2.5. After hydration, or hydration and reheating, the size of the MPs was determined with static light scattering (Mastersizer 2000, Malvern Instruments). Also the size of the unhydrated MPs (i.e., MPs obtained after oven drying and milling and referred to as dry MPs) was determined. Therefore, a 10% w/w dispersion was made with isopropanol, a solvent in which the MPs do not swell. During the measurements, water was used as a carrier fluid for the hydrated MPs and isopropanol for the dry MPs. The rotation speed of the vessel of the carrier fluid was set at 1200 rpm. Refractive indices of 1.545 (MPs), 1.33 (water), and 1.39 (isopropanol) and an absorption of 0.001 (MPs) were used for the calculations €m, Nielsen, Windhab, & Hermansson, 1999). The (Walkenstro measurements were done for three dispersions per type of MP; the size distribution was measured in quintuplicate. The size of the dry MPs was measured in triplicate. From the measurements, the average particle diameter d4,3 (volume mean diameter) was calculated assuming that the MPs were spherical. The average diameter d4,3 was used to calculate the average volume of the dry MPs and the volume increase when hydrated, or hydrated and reheated. The amount of water inside and between the MPs was calculated from the WBC of the pellets, assuming that the proteins did not dissolve and that the density of dry MPs (Purwanti, van der Veen, van der Goot, & Boom, 2013) was 1.3 g cm3 and the density of water was 1 g cm3. 2.2.8. 1H relaxometry 1 H relaxometry was performed with a Maran Ultra NMR spectrometer (at 30.7 MHz proton resonance frequency; Resonance Instruments Ltd., Witney, United Kingdom). T2 relaxation decay curves were recorded by means of a standard Carr-PurcellMeiboom-Gill (CPMG) pulse sequence. The CPMG decay train consisted of 16,384 echoes with an echo time of 1 ms. Experiments were averaged over 16 scans with a repetition time of 30 s. T2 spectra of relative intensity as a function of T2 were determined by numerical inverse Laplace transformation of the data as implemented in CONTIN (Provencher, 1982). The regularization parameter in CONTIN was set to the same, rather conservative value, for all spectra to facilitate comparison of the results. This is justified by the fact that the signal-to-noise ratio was comparable
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
for all measurements. T2 spectra were subsequently analyzed by determining peak positions (the top of the peaks was used to determine the T2 values of the various water fractions in the samples) and areas to determine relative amounts using IDL (ITT Visual Information Solution, Boulder, CO). For decays with a clear mono-exponential character, the LevenbergeMarquardt non-linear least squares algorithm implemented in SPLMOD (Provencher & Vogel, 1983) was used to analyze the data. These decays were measured to obtain T2 values for calibration purposes. The rationale for using SPLMOD instead of CONTIN in those cases is that SPLMOD fits both the amplitude and T2 relaxation time of a discrete sum of exponentials. CONTIN, on the other hand, determines amplitudes given a distribution of fixed T2 values. In the latter, an extra unnecessary step is needed to obtain average T2 values. Every sample was measured once because extra experiments performed on a number of samples showed good reproducibility. This allowed us to measure a broader range of samples. 2.2.8.1. 1H relaxometry of whey protein microparticle pellets. Pellets were made as described in Section 2.2.5. After decanting the supernatant from the pellet, the bottoms of the Eppendorf tubes were cut off so the samples were disturbed as little as possible before analysis (scooping the pellet from the tube affected the results). The tips of the tubes were used for further analysis. The resulting transverse decay curves were all multi-exponential and analyzed with CONTIN. To estimate the absolute amount of water present inside and between the MPs, the amount of water present according to the WBC of the pellets was divided over those two fractions according to the ratio of these areas in the CONTIN T2 spectra. Using the T2 values for the top of the two fractions, the dry matter content inside and between all the various MP pellets was calculated from the standard calibration curves (Supporting information). The calibration curve obtained for solutions and gels made at pH 5.8 (Supporting information) was used only for the MPs made at pH 5.8. From the amount of dry matter that was present between the MPs the percentage of dry matter dissolved from the MPs was calculated. Though, if the mass balance was used the solubility would be lower or larger than calculated via the WBC, T2 values and the calibration curves. The relative area under the peaks was determined by dividing either the area of the peak attributed to water inside the MPs or between the MPs over the sum of the total area.
173
obtained were all assumed to be mono-exponential and analyzed with SPLMOD. 2.2.8.4. 1H relaxometry of unheated and heated whey protein isolate solutions. Unheated and heated WPI solutions, which were made according to Section 2.2.4, were analyzed using TD NMR by pipetting 125 mL of the solution into the NMR tubes. The resulting curves were analyzed with SPLMOD because of the mono-exponential character of the decay curves. 2.2.8.5. 1H relaxometry of whey protein microparticles hydrated in desiccators. After hydration of the MPs in the desiccators (Section 2.2.6), the hydrated MPs were placed in an NMR tube and analyzed. The curves obtained were assumed to be mono-exponential and analyzed with SPLMOD. 2.2.9. Dry matter content The dry matter content of the gels, solutions, and MPs was determined by drying the samples in an oven at 105 C for 24 h. 3. Results and discussion The following sections describe the steps that were taken to gain inside into the water-binding capacity (WBC) of whey protein microparticles (MPs). First, the WBC of the MPs was determined, and then the value was compared to the swelling of the MPs as calculated from the light scattering data. Subsequently, the pellets were analyzed with time domain nuclear magnetic resonance (TD NMR). To be able to further interpret these TD NMR results, swollen MPs were mimicked by MPs subjected to certain relative humidities (RHs) and whey protein gels, and also analyzed with TD NMR. In addition, the dry matter and water concentrations in both water fractions were calculated and compared with the light scattering results. Finally, the water fractions obtained were assigned to the two water domains in the MP pellets (water inside and between the MPs) and the contribution of both water domains to the WBC of the pellet was investigated.
2.2.8.2. 1H relaxometry of supernatants. The spectra of the supernatant of standard MPs, unwashed MPs incubated in 40 mM DTT, and MPs made at pH 5.8 were obtained by making pellets and supernatants first as described in Section 2.2.5. After decanting the supernatant from the pellet, 125 mL was added to the NMR tube and analyzed. The resulting transverse decay curves were assumed to be mono-exponential and analyzed with SPLMOD. The SPLMOD results for the supernatants were used to calculate the dry matter content of the supernatants. For that, only the solution part of the calibration curve was used (Supporting information). For the supernatant of the standard MPs and the unwashed MPs incubated in 40 mM DTT, the calibration curve of the standard solutions was used; for the MPs made at pH 5.8, the curve for the solutions made at pH 5.8 was used. 2.2.8.3. 1H relaxometry of macro gels. After the formation of the macro gels, as described in Section 2.2.3, the gels were cut with a biopsy punch into small cylinders and analyzed. The curves
Fig. 1. The various water domains formed after centrifuging whey protein microparticles (MPs) that were hydrated in an excess amount of water: (A) supernatant, (B) water between the MPs, and (C) water inside the MPs.
174
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
3.1. Analysis of the swelling of whey protein microparticles in a pellet with light scattering To determine the WBC of MPs, the MPs were hydrated in excess water after which the dispersions were centrifuged to obtain a pellet and supernatant (Fig. 1). The pellet, containing swollen MPs and water between the MPs, was weighed to determine the WBC, assuming that the amount of water between the MPs was negligible as an initial hypothesis. However, the importance of water between the MPs is investigated in this study. That is why the term WBC-P (WBC of the pellet) is introduced to distinguish the WBC of the MPs themselves (WBC-MPs) from the WBC of the pellet that includes water between the MPs as well. The WBC-P of the unwashed MPs treated with 40 mM DTT and the MPs made at pH 5.8 differed most from the WBC-P of standard MPs from all tested MPs. Therefore, those three types of MPs were analyzed in more detail. The unwashed MPs treated with 40 mM DTT had a much larger WBC-P than the standard MPs, whereas MPs made at pH 5.8 had a much lower WBC-P (Table 1). However, it is not clear whether this WBC-P can be translated to the WBC-MPs directly or that the water between the MPs contributed to the WBC-P as well. A first route to estimate the relative importance of water inside and between the MPs is to consider the swelling of the MPs. Calculations based on the average particle diameter d4,3 (volume mean diameter) of these MPs (see Section 2.2.7) suggested that the unheated MPs treated with DTT contained more water than the standard MPs, whereas both the unheated and reheated MPs made at pH 5.8 contained less water than the standard MPs (Table 1). Although these results are indicative (among others, because the particles were not spherical (Peters et al., 2015)), they strongly suggest that swelling of the MPs only cannot explain the WBC-P. In all cases, the water content inside the swollen MPs was significantly lower than the total amount of water present in the pellets (WBC-P). This implies that a certain amount of water was present between the MPs since the pellet consists of water inside and between the MPs. The WBC-MPs is therefore most probably lower than the WBC-P.
Fig. 2. The CONTIN T2 spectra of (A) pellets and (B) supernatant formed during the water-binding experiment with standard whey protein microparticles (ST MPs) (solid line), unwashed MPs treated with 40 mM DTT (DTT MPs) (dashed line), and MPs made at pH 5.8 (pH MPs) (dotted line).
fraction appearing around T2 ¼ 1 s seems to be related to water that is not bound within the pellet. The T2 value of both the supernatant and expelled water fraction was lower than the T2 value of free water, which was measured to be around 2.4 to 2.7 s, suggesting the presence of proteins in the supernatant and expelled water. The other two fractions are therefore ascribed to water present inside the pellets. The T2 values of the fractions with the smallest transverse relaxation time (T2,1) were different for the three samples. A smaller WBC corresponded to a smaller T2,1, suggesting a larger dry matter content in that water fraction. The second fraction (T2,2) had a T2 value around 0.1 s, which was similar for all three pellets. The ratio of the areas with T2,1 and T2,2 differed between the pellets as depicted in Fig. 2A. The unwashed MPs treated with 40 mM DTT had a larger area for the T2,2 fraction than for the T2,1
3.2. Analysis of pellets of whey protein microparticles with time domain nuclear magnetic resonance Time domain nuclear magnetic resonance (TD NMR) was used to further analyze the water distribution in the pellets. CONTIN transverse relaxation time (T2) spectra for the pellets of standard MPs, unwashed MPs treated with 40 mM DTT, and MPs made at pH 5.8 showed two main peaks and one small peak around T2 ¼ 1 s (Fig. 2A), indicating the presence of three different water fractions. The fraction with the largest transverse relaxation time (T2,3) was assigned to expelled water that was present after cutting the tube with the pellet. This assignment was done based on the fact that the supernatant obtained after centrifugation (domain A in Fig. 1) also showed a T2 fraction around 1 s (Fig. 2B). Therefore, the water
Table 1 The water-binding capacity (WBC) of the unheated and reheated pellets (g water/g dry matter) of standard (ST) whey protein microparticles (MPs), unwashed MPs treated with 40 mM dithiothreitol (DTT MPs) and MPs made at pH 5.8 (pH MPs), the average size of the dry, hydrated, and hydrated and reheated MPs (mm) (d4,3), and the calculated water content (g water/g dry matter) and dry matter content (%) inside the unheated and reheated swollen MPs. WBC pellets (g water/g dry matter)
Average size MPs (mm) (d4,3)
Unheated
Reheated
Dry
5.4 (0.1) 7.9 (0.0) 2.4 (0.1)
79 (2) 110 (2) 78 (2) 120 (3) 74 (3) 99 (4)
ST MPs 3.7 (0.2) DTT MPs 6.9 (0.2) pH MPs 2.1 (0.2)
Calculated water content inside swollen MPs (g Calculated dry matter content of water/g dry matter) swollen MPs (%)
Hydrated Hydrated and reheated Unheated
The standard deviations are given in parentheses.
127 (2) 114 (3) 98 (2)
1.3 2.0 1.1
Reheated
Unheated
Reheated
2.4 1.6 1.0
43 33 48
29 38 50
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
175
fraction; the areas for these fractions were nearly the same for the MPs made at pH 5.8. For further insight into the WBC-P, the pellets of all the other MPs (MPs treated with DTT washed and unwashed, MPs treated with DTT and N-ethylmaleimide, MPs treated with genipin, and MPs treated with transglutaminase (Tgase) by addition of Tgase to the WPI solution or the wet gel particles), both unheated and reheated, were analyzed with TD NMR. The CONTIN T2 spectra of all MP pellets showed at least the two water fractions T2,1 and T2,2. Fig. 3A shows that T2,1 increased from WBC-P 2 to 5.5 g water/g dry matter, after which it tends to deviate from the trend. No clear relationship was found between T2,2 and the WBC-P (Fig. 3B), because all pellets had a T2,2 value around 0.12 (±0.02). The relative area2,2 was almost linearly related to the WBC-P (Fig. 3C). Therefore, these results suggest that an increase in WBC-P is caused by both an increase in the amount of water in water fraction T2,1 (T2,1 increased with an increase in the WBC implying an increase in the water content) and in water fraction T2,2 (the relative area2,2 increased with an increase in WBC). 3.3. Assignment of water fractions obtained with time domain nuclear magnetic resonance to the water domains in whey protein microparticle pellets It is likely that T2,1 and T2,2 can be assigned to the two water domains within the pellet: the swollen MPs and water between the MPs. The exchange rate of protons and the water/protein ratio are probably larger inside the MPs (domain C in Fig. 1) than between the MPs (domain B in Fig. 1). Therefore, it can be hypothesized that T2,1 represents water inside the MPs and T2,2 the water between the MPs. However, from the TD NMR results of the pellets only it is not clear if this assignment can be done instantly, or that water in the pellet was divided over two water peaks for other reasons. Therefore, it would be the best to analyze swollen MPs and water between the MPs separately. Since it would be hard to separate both fractions from the pellet, here it was tried to mimic swollen MPs to get to know if T2,1 could be assigned to this water domain. To get hydrated MPs without water between the MPs, MPs were placed in a desiccator with controlled RHs for 12 days. A TD NMR analysis of those MPs resulted in spectra with mainly one fraction, as illustrated in Fig. 4A for the standard MPs. Smaller T2 values were found when less water was present. The T2 value of the MPs at RH 95% was still one order smaller than the T2,1 value of the pellet spectra. This relatively large difference corresponds to the isotherm of these MPs, which shows a strong increase in moisture content from aw 0.95 to 1 (Peters et al., 2015). In addition, whey protein isolate (WPI) macro gels were analyzed with TD NMR, because macro gels were considered as a model material for MPs without surrounding water but with a large moisture content. Fig. 4B shows that all the spectra for WPI macro gels contained one main fraction in the same range as the T2,1 fraction of the pellets (Fig. 2A). Mono-exponential behavior for WPI and whey protein concentrate gels with a protein concentration of 6 to 36% (w/w) has also been found by others (Colsenet, Mariette, & Cambert, 2005; Lambelet, Berrocal, Desarzens, Froehlicher, & Ducret, 1988; Oztop, Rosenberg, Rosenberg, McCarthy, McCarty, 2010). As expected, the T2 value of the gels decreased when the WPI concentration of the gel increased (Fig. 4B), because an increase in the protein content increases the total interaction with water and reduces the transverse relaxation time (Oakes, 1976b). The spectra of the WPI macro gels and MPs hydrated in a desiccator suggest both that one fraction (T2,1) within the pellet spectra is associated with water inside the MPs, and that therefore T2,2 can be assigned to water between the MPs.
Fig. 3. Relationship between the water-binding capacity (WBC-P) (g water/g dry matter) of the unheated (filled diamond) and reheated (open diamond) pellets and (A) T2,1 (s), (B) T2.2 (s), and (C) the relative area2,2 (-).
Another method to check whether T2,1 can be assigned to water inside the MPs and T2,2 to water between the MPs is by calculating the water and dry matter concentrations within those domains and comparing those concentrations to the light scattering data. Therefore, a standard calibration curve relating the dry matter content to the transverse relaxation rate R2 (R2 ¼ 1/T2) was drawn (Supporting information) (Colsenet et al., 2005; Lambelet et al., 1988; Oakes, 1976a, 1976b). A separate calibration curve was made for the MPs made at pH 5.8, because those samples showed a different relationship between R2 and the dry matter content (Supporting information). The dry matter content in the supernatant of the standard MPs and MPs made at pH 5.8 could be estimated well (Table 2), indicating the validity of the calibration
176
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
Fig. 5. Relationship between the water-binding capacity of the pellets (WBC-P) (g water/g dry matter) and the amount of water (g water/g dry matter) present inside (filled diamond) and between (open diamond) the whey protein microparticles (MPs).
Fig. 4. The CONTIN T2 spectra of (A) the standard microparticles incubated in a desiccator with RH 25% (dotted line), 46% (dashed line), and 95% (solid line); and (B) whey protein isolate (WPI) gels with a WPI concentration (w/w) of 19% (solid line), 30% (dashed line), and 39% (dotted line).
curves. The supernatant values obtained for the MPs treated with 40 mM DTT show some discrepancy between the measured and calculated dry matter content. These samples seem to be a special case as they also deviate from the trend in Fig. 3A. The dry matter content inside the MPs estimated from the TD NMR data and the calibration curves (Table 2) compares well with the light scattering data for the standard MPs, the unwashed MPs incubated in 40 mM DTT and the MPs made at pH 5.8 (Table 1). So, it was concluded that T2,1 can be assigned to water inside the MPs and T2,2 to water between the MPs. 3.4. Contribution of water inside and between the whey protein microparticles to the water-binding capacity of the pellets The next step is to estimate the contribution of water inside and between the MPs to the WBC-P. Therefore, the total amount of water present in the pellets is divided into two water fractions according to the ratio between the areas under the curve of those two fractions. This division shows that the amount of water present inside the MPs increased slightly in the WBC-P ranging from 2 to 5.5 g water/g dry matter (Fig. 5). It seems that the increase in T2,1 with increasing WBC-P was less important than the decrease in the relative area2,1 with increasing WBC-P. The increase in the WBC-P appears to be mainly related to an increase in the amount of water between the MPs. A consequence of that reasoning is that the contribution of water between the MPs cannot be neglected in the WBC-P, and that the WBC-MPs cannot be determined via the WBC-P.
The contribution of water between the MPs to the WBC-P seems to be large. At a WBC of 5.5 g water/g dry matter 80% of the water was present between the MPs, which implies that water-binding in this fraction in the pellet probably consisted of more than water inclusion by the MPs only. A remarkable observation for all MP pellets is that they had nearly the same T2,2 (Fig. 3B), which suggests a similar dry matter content between the MPs. Consequently, it seems that dissolved proteins originating from the MPs are a determining factor for the WBC-P. If it is assumed that the T2 of water is caused by dissolved proteins only, the T2,2 values suggest the presence of around 11 to 15% dry matter between the MPs. Although a certain amount of dry matter was expected due to the fact that the supernatant contains proteins as well (Fig. 2B and Table 2), this high dry matter content was unexpected. The suggested higher dry matter content between the MPs could be related to small highly swollen protein aggregates that were dragged down on centrifugation. The protein content suggested by the T2,2 value could also explain the binding of water between the MPs, because this concentration might have led to the formation of a (particle) gel, or an increased viscosity at least, upon centrifugation. Without such effect, more water would be expelled during centrifugation. Rough calculations show that the volume fraction of standard MPs in a pellet is around 0.5 only, which is clearly lower than the maximum packing density of randomly packed monodisperse spheres (around 0.65) while polydisperse systems could have an even higher packing density (Walstra, 2003). Nevertheless, the possibility that T2,2 was influenced by other interactions and/or capillary pressures cannot be excluded. In that case, the actual protein concentration is somewhat lower than indicated by T2,2, but still much larger than the protein content in the supernatant. Fig. 5 also shows the different behavior of the unwashed DTT treated particles (the particles with a WBC-P larger than 5.5 g water/g protein), compared to the other MPs. The differences are probably due to the fact that no clear distinction can be made between water inside and between those particular MPs due to overlapping water peaks in the TD NMR spectra. This seems likely, because the area of T2,1 for the unwashed MPs incubated in 40 mM DTT was small compared with the area of T2,1 for the other MPs (Fig. 2A). It may be that those MPs incubated in 40 mM DTT partly solubilized, leading to a nearly equal dry matter and water content inside and between the MPs. After the division of water according to the ratio of the areas, the WBC-P was calculated from the concentrations obtained via calibration curves. A comparison between the WBC-P calculated from
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
177
Table 2 The measured dry matter content (%) (standard deviation) and calculated dry matter content (%) with the calibration curves in the supernatants of the unheated and reheated standard (ST) whey protein microparticles (MPs), unwashed MPs treated with 40 mM DTT (DTT MPs), and MPs made at pH 5.8 (pH MPs), and the calculated dry matter content (%) and solubility of these MPs (%) in the unheated and reheated pellets. Supernatant
ST MPs DTT MPs pH MPs
Pellet
Measured dry matter content (%)
Calculated dry matter content (%)
Calculated dry matter content (%)
Unheated
Unheated
Unheated
0.6 (0.1) 2.2 (0.5) 0.5 (0.1)
Reheated
0.6 (0.1) 3.2 (0.5) 0.5 (0.1)
0.7 1.2 0.4
Reheated
0.9 2.1 0.4
the TD NMR data and the measured WBC-P shows that the WBC-P measured from the TD NMR data was mostly larger than the measured WBC-P (Fig. 6). This is probably caused by the way the WBC-P was determined when using the pellet weight after centrifugation. This method does not consider the solubilization of dry matter, and therefore it can overestimate the dry matter content and consequently underestimate the WBC. A 10% smaller dry matter content can explain the difference between the measured WBC-P and the calculated WBC-P from the TD NMR data as shown in Fig. 6 with the dotted line. With TD NMR, on the other hand, the pellet obtained after centrifugation was used for analysis and the water concentration was obtained from measuring 1H protons. This has probably resulted in more accurate values for the dry matter and water concentrations within both water domains in the pellets and a more accurate WBC-P. Therefore, it can be concluded that with TD NMR a distinction between water present inside and between the MPs in a pellet can be made. In addition, TD NMR is a more exact method to determine the WBC-P and the dry matter and water concentrations in the water fractions of the pellet than the use of the pellet weight obtained after centrifugation of a MP dispersion. 4. Conclusions The water-binding capacity (WBC) of protein samples is frequently determined from the pellet weight obtained after centrifuging a dispersion of this sample, assuming that water
Solubility of MPs (%)
Reheated
Between MPs
Inside MPs
Between MPs
Inside MPs
12 12 12
35 31 59
9 12 8
25 27 60
Unheated
Reheated
33 89 23
39 87 15
between the sample particles can be neglected. However, we have demonstrated in this paper that the amount of water between the particles cannot be neglected. For pellets made of whey protein microparticle (MP) dispersions, with time domain nuclear magnetic resonance (TD NMR) a distinction was made between water inside and between the MPs. It was shown that an increase in WBC of the pellet (WBC-P) from 2 to 5.5 g water/g dry matter was partly caused by a slight increase in swelling of the MPs, but to a larger extent by an increase in the amount of water between the MPs. The latter could account for as much as 80% of the water being kept in a pellet with a WBC of 5.5 g water/g dry matter. Therefore, it is concluded that the WBC-P is not a measure for the WBC of MPs (WBC-MPs); the outcomes of the centrifugation method should be interpreted cautiously when used to determine the WBC-MPs. We hypothesize that this relatively large amount of water between the MPs could be present because small highly swollen protein structures co-sedimented between the MPs during centrifugation. The presence of such structures can also explain the remarkable result of a nearly constant T2 value at all WBC-Ps, which suggest the presence of 11 to 15% of dry matter in the water fraction of water between the MPs. The results presented in this paper demonstrate that TD NMR is a better method to determine the WBC-MP, the amount of water between the MPs and the overall WBC-P than using the pellet weight after centrifugation. We therefore conclude that TD NMR used in the manner proposed is a useful additional tool to understand WBC of a dispersed particle system. Acknowledgments The authors would like to thank FrieslandCampina and NanoNextNL, a consortium of the government of the Netherlands and 130 partners, for their financial support of this research. In addition, we would like to thank Antoinette Toebes of FrieslandCampina for help with the TD NMR experiments and reading of the manuscript. Appendix. Supporting information Supporting information related to this article can be found at http://dx.doi.org/10.1016/j.foodhyd.2015.09.031. References
Fig. 6. The measured water-binding capacity of the pellets (WBC-P) (g water/g dry matter) and the WBC-P calculated from time domain nuclear magnetic resonance (TD NMR data) of the unheated (filled diamond) and reheated (open diamond) MPs. The WBC-P was calculated from the ratio between the areas of water inside (T2,1) and outside (T2,2) the MPs from the CONTIN T2 spectra and the protein concentrations obtained from the calibrations curves. The solid line shows the relation if the measured WBC-P was the same as the TD NMR WBC-P. The dashed line shows the relation in case the dry matter content of the pellets was actually 10% higher than used to calculate the measured WBC-P.
Ahmed, J., Al-Jassar, S., & Thomas, L. A. (2015). Comparison in rheological, thermal, and structural properties between Indian Basmati and Egyptian Giza rice flour dispersions as influenced by particle size. Food Hydrocolloids, 48, 72e83. Berghout, J. A. M., Boom, R. M., & van der Goot, A. J. (2014). The potential of aqueous fraction of lupin seeds for high-protein foods. Food Chemistry, 159, 64e70. Bertram, H. C., Andersen, H. J., & Karlsson, A. H. (2001). Comparative study of lowfield NMR relaxation measurements and two traditional methods in the determination of water holding capacity of pork. Meat Science, 57(2), 125e132. Bertram, H. C., Dønstrup, S., Karlsson, A. H., & Andersen, H. J. (2002). Continuous distribution analysis of T2 relaxation in meatdan approach in the determination of water-holding capacity. Meat Science, 60(3), 279e285.
178
J.P.C.M. Peters et al. / Food Hydrocolloids 54 (2016) 170e178
Brøndum, J., Munck, L., Henckel, P., Karlsson, A., Tornberg, E., & Engelsen, S. B. (2000). Prediction of water-holding capacity and composition of porcine meat by comparative spectroscopy. Meat Science, 55(2), 177e185. Colsenet, R., Mariette, F., & Cambert, M. (2005). NMR relaxation and water selfdiffusion studies in whey protein solutions and gels. Journal of Agricultural and Food Chemistry, 53(17), 6784e6790. van Duynhoven, J., Voda, A., Witek, M., & Van As, H. (2010). Time-domain NMR applied to food products. In A. W. Graham (Ed.), Annual reports on NMR spectroscopy (Vol. 69, pp. 145e197). Academic Press. Fennema, O. R. (1996). Food chemistry (3rd ed., pp. 366e370). Taylor & Francis. Goetz, J., & Koehler, P. (2005). Study of the thermal denaturation of selected proteins of whey and egg by low resolution NMR. LWT e Food Science and Technology, 38(5), 501e512. € tz, J., Noll, M., Wolfschoon, A., Eibel, H., & Weisser, H. (2004). Hinrichs, R., Go Characterisation of the water-holding capacity of fresh cheese samples by means of low resolution nuclear magnetic resonance. Food Research International, 37(7), 667e676. €tz, J., & Weisser, H. (2003). Water-holding capacity and structure of Hinrichs, R., Go hydrocolloid-gels, WPC-gels and yogurts characterised by means of NMR. Food Chemistry, 82(1), 155e160. Jain, A., Prakash, M., & Radha, C. (2015). Extraction and evaluation of functional properties of groundnut protein concentrate. Journal of Food Science and Technology. http://dx.doi.org/10.1007/s13197-015-1758-7. Kneifel, W., Paquin, P., Abert, T., & Richard, J. P. (1991). Water-holding capacity of proteins with special regard to milk proteins and methodological aspectsda review. Journal of Dairy Science, 74(7), 2027e2041. Lambelet, P., Berrocal, R., Desarzens, C., Froehlicher, I., & Ducret, F. (1988). Pulsed low-resolution NMR investigations of protein sols and gels. Journal of Food Science, 53(3), 943e946. Mariette, F. (2006). NMR relaxation of dairy products. In G. A. Webb (Ed.), Modern magnetic resonance (pp. 1697e1701). Springer. Mitchell, J., Gladden, L. F., Chandrasekera, T. C., & Fordham, E. J. (2014). Low-field permanent magnets for industrial process and quality control. Progress in Nuclear Magnetic Resonance Spectroscopy, 76, 1e60. Oakes, J. (1976a). Protein hydration. Nuclear magnetic resonance relaxation studies of the state of water in native bovine serum albumin solutions. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
72, 216e227. Oakes, J. (1976b). Thermally denatured proteins. Nuclear magnetic resonance, binding isotherm and chemical modification studies of thermally denatured bovine serum albumin. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 72, 228e237. Oztop, M. H., Rosenberg, M., Rosenberg, Y., McCarthy, K. L., & McCarthy, M. J. (2010). Magnetic resonance imaging (MRI) and relaxation spectrum analysis as methods to investigate swelling in whey protein gels. Journal of Food Science, 75(8), E508eE515. Peters, J. P. C. M., Luyten, H., Alting, A. C., Boom, R. M., & van der Goot, A. J. (2015). Effect of crosslink density on the water-binding capacity of whey protein microparticles. Food Hydrocolloids, 44, 277e284. Provencher, S. W. (1982). A constrained regularization method for inverting data represented by linear algebraic or integral equations. Computer Physics Communications, 27(3), 213e227. Provencher, S. W., & Vogel, R. H. (1983). Regularization techniques for inverse problems in molecular biology. In P. Deuflhard, & E. Hairer (Eds.), Numerical treatment of inverse problems in differential and integral equations (Vol 2, pp. 304e319). Boston: Birkh€ auser. Purwanti, N., van der Veen, E., van der Goot, A. J., & Boom, R. (2013). Stiffening in gels containing whey protein isolate. International Dairy Journal, 28(2), 62e69. Straadt, I. K., Rasmussen, M., Andersen, H. J., & Bertram, H. C. (2007). Aging-induced changes in microstructure and water distribution in fresh and cooked pork in relation to water-holding capacity and cooking loss e a combined confocal laser scanning microscopy (CLSM) and low-field nuclear magnetic resonance relaxation study. Meat Science, 75(4), 687e695. Sun, Q., & Xiong, C. S. L. (2014). Functional and pasting properties of pea starch and peanut protein isolate blends. Carbohydrate Polymer, 101, 1134e1139. €m, P., Nielsen, M., Windhab, E., & Hermansson, A. M. (1999). Effects of Walkenstro flow behaviour on the aggregation of whey protein suspensions, pure or mixed with xanthan. Journal of Food Engineering, 42(1), 15e26. Walstra, P. (2003). Physical chemistry of foods (p. 99). Marcel Dekker, Inc. Yu, J., Ahmedna, M., & Goktepe, I. (2007). Peanut protein concentrate: production and functional properties as affected by processing. Food Chemistry, 103(1), 121e129. Zayas, J. F. (1997). Functionality of proteins in food (pp. 76e133). Springer Berlin Heidelberg.